Table 3

Demographics, clinical and tumor characteristics, treatment modalities associated with overall survival by Cox univariate analysis in 87 patients
n (%) or mean ± SD Hazard ratio (95% confidence interval) p value
Age 72.04 ± 8.427 1.044(0.998-1.093) 0.063
Grouped by age
<73 37(44.58%) 1
≥73 46(55.42%) 1.521(0.696-3.326) 0.293
Race
Non-Asian 24(27.59%) 1
Asian 63(72.41%) 0.291(0.140-0.606) 0.001
Clinical symptoms
Hydronephrosis
No 70(80.4%) 1
Yes 17(19.54%) 0.904(0.389-2.104) 0.786
Dysuria
No 69(79.31%) 1
Yes 18(20.69%) 1.939(0.783-4.805) 0.815
Dyschezia
No 27(31.03%) 1
Yes 60(68.97%) 1.193(0.557-2.556) 0.650
bowel obstruction
No 80(91.9%) 1
Yes 7(8.1%) 0.821(0.292-2.310) 0.709
Rectum bleeding
No 67(77.01%) 1
Yes 20(22.99%) 2.145(1.023-4.498) 0.043
The Time to development of RI
Synchronous presentation 58(66.67%) 1
Metachronous presentation 29(33.33%) 14.339(5.719-35.952) 0.000
Tumor characteristics at diagnosis of rectal involvement
Pathological grade of RI
well or moderately differentiated 12(17.14%) 1
poorly differentiated 58(82.86%) 2.048(0.474-8.853) 0.337
Histology of prostate cancer RI
adenocarcinoma 81(93.10%) 1
Mixed type 6(6.90%) 0.356(0.048-2.615) 0.310
CEA
<5 ng/mL 26(83.87%) 1
≥5 ng/mL 5(16.13%) 2.795(0.510-15.331) 0.237
CA199
<37 u/mL 23(85.19%) 1
≥37 u/mL 4(14.81%) 2.150(0.223-20.768) 0.508
PSA 464.74 ± 1189.28 1.000(0.998-1.000) 0.499
PSA
<92.41 u/mL 28(52.83%) 1
≥92.41 u/mL 25(47.17%) 0.136(0.153-1.248) 0.122
Type of rectal involvement
Type I 9(13.04%) 1
Type II 40(57.97%) 0.494(0.151-1.615) 0.244
Type III 11(15.94%) 1.163(0.311-4.346) 0.822
Type IV 9(13.04%) 1.653(0.433-6.304) 0.462
Hormonal status
No prior hormonal therapy 61(70.11%) 1
Prior hormonal therapy 26(29.89%) 14.658(6.042-35.561) 0.000
Tumor extent at diagnosis of RI
rectum-only 24(35.82%) 1
Extrarectal metastases 43(64.18%) 2.618(0.984-6.961) 0.054

RI: rectal involvement.

Type of RI was defined according to the criteria described in the literature [11,12].

CEA and CA199 were analyzed as categorical variables using normal value as cut off; Age and PSA were analyzed as categorical variables using median value as cut off.

Wang et al.

Wang et al. Diagnostic Pathology 2014 9:35   doi:10.1186/1746-1596-9-35

Open Data